Does bullying cause emotional problems? A prospective study of young teenagersBMJ 2001; 323 doi: http://dx.doi.org/10.1136/bmj.323.7311.480 (Published 01 September 2001) Cite this as: BMJ 2001;323:480
- Lyndal Bond, head, research unit ()a,
- John B Carlin, directorb,
- Lyndal Thomas, PhD scholarb,
- Kerryn Rubin, medical studenta,
- George Patton, directora
- a Centre for Adolescent Health, Royal Children's Hospital, Parkville 3052, Victoria, Australia
- b Clinical Epidemiology and Biostatistics Unit, Murdoch Children's Research Institute and University of Melbourne Department of Paediatrics, Royal Children's Hospital
- Correspondence to: L Bond
- Accepted 21 May 2001
Objectives: To establish the relation between recurrent peer victimisation and onset of self reported symptoms of anxiety or depression in the early teen years.
Design: Cohort study over two years.
Setting: Secondary schools in Victoria, Australia.
Participants: 2680 students surveyed twice in year 8 (aged 13 years) and once in year 9.
Main outcome measures: Self reported symptoms of anxiety or depression were assessed by using the computerised version of the revised clinical interview schedule. Incident cases were students scoring ≥12 in year 9 but not previously. Prior victimisation was defined as having been bullied at either or both survey times in year 8.
Results: Prevalence of victimisation at the second survey point in year 8 was 51% (95% confidence interval 49% to 54%), and prevalence of self reported symptoms of anxiety or depression was 18% (16% to 20%). The incidence of self reported symptoms of anxiety or depression in year 9 (7%) was significantly associated with victimisation reported either once (odds ratio 1.94, 1.1 to 3.3) or twice (2.30, 1.2 to 4.3) in year 8. After adjustment for availability of social relations and for sociodemographic factors, recurrent victimisation remained predictive of self reported symptoms of anxiety or depression for girls (2.60, 1.2 to 5.5) but not for boys (1.36, 0.6 to 3.0). Newly reported victimisation in year 9 was not significantly associated with prior self report of symptoms of anxiety or depression (1.48, 0.4 to 6.0).
Conclusion: A history of victimisation and poor social relationships predicts the onset of emotional problems in adolescents. Previous recurrent emotional problems are not significantly related to future victimisation. These findings have implications for how seriously the occurrence of victimisation is treated and for the focus of interventions aimed at addressing mental health issues in adolescents.
What is already known on this topic
What is already known on this topic Being bullied is a common experience for many young people
Victimisation is related to depression and, to a lesser extent, anxiety, loneliness, and general self esteem
Debate remains as to whether victimisation precedes the onset of emotional problems or whether young people with emotional problems “invite” victimisation
What this study adds
What this study adds A history of victimisation predicts the onset of anxiety or depression, especially in adolescent girls
Previous recurrent emotional problems are not significantly related to future victimisation
Reduction in bullying in schools could have a substantial impact on the emotional wellbeing of young people
Bullying occurs in all schools, but its relevance to health and wellbeing is uncertain.1-3 On the one hand it can be considered a common and normal developmental experience; alternatively, it can be considered an important cause of stress and of physical and emotional problems.4-6 A meta-analysis of studies investigating the relation between victimisation and psychosocial maladjustment found a stronger association with measures of depression than with anxiety, loneliness, or general self esteem.1
Unfortunately, the cross sectional design of most studies precludes inferences about causality. The few available prospective studies have generally focused on primary school children before the early increase in depression in adolescence,7 with the principal outcomes being school maladjustment, loneliness, and depression.8-10 One small longitudinal study of adolescents found that high levels of victimisation predicted poor physical health for boys and girls and poor mental health for girls.11 Olweus found that boys victimised between the ages of 12 and 16 had increased levels of depression as young adults; however, no adjustment was made for previous mental health states in this study.5
We carried out a prospective study of secondary school students. The data derive from three waves of data collected from students involved in a randomised controlled trial of a school based intervention to promote the emotional wellbeing of young people.12 Intervention effects in the trial are not the main focus of this paper. Data were collected at the beginning and end of year 8 (second year of secondary school, mean age 13 years) and 12 months later (end of year 9). Our aim was to use these prospective data to examine the relation between a history of victimisation (in year 8) and the incidence of self reported symptoms of anxiety or depression in year 9.
A cluster randomised controlled design was used for the allocation of education districts to intervention or control status. In metropolitan Melbourne, 12 districts were sampled with a probability proportional to the number of secondary schools (including government, independent, and Catholic schools) and were randomly allocated to intervention or control status. We used simple random sampling to select 12 schools from the “intervention” districts and 12 from the “control” districts. Six country schools were randomly drawn from two regional districts. Twenty six (12 intervention and 14 control) schools agreed to participate.
Students completed a self administered questionnaire at school using laptop computers provided by the research team. Questionnaires took approximately 40 minutes to complete. Absent students were surveyed at school at a later date or by telephone.
Ethics approval was granted by the Royal Children's Hospital ethics in human research committee, the Victorian Department of Education, Employment and Training, and the Catholic Education Office. Student participation was voluntary, with written parental consent required.
Participants were classified as victimised if they answered yes to items addressing four types of recent victimisation: being teased, having rumours spread about them, being deliberately excluded, or experiencing physical threats or violence. Respondents were classified on the basis of self report in year 8 as having experienced recurrent victimisation if they reported having been bullied at both times in year 8 (waves 1 and 2).
Mental health status
Mental health status was evaluated with a computerised version of the revised clinical interview schedule, a structured psychiatric interview for non-clinical populations. 13 14 The schedule comprises 14 subscales. It has been used as a criterion measure for the definition of caseness in teenagers,15 and it has an ease of reading suitable for young adolescents (Fleisch reading ease 78.5). A score of ≥12 provides a criterion measure of minor psychiatric morbidity at which a general practitioner might be concerned.14
An incident case with self reported symptoms of anxiety or depression was defined as someone who scored <12 on the interview schedule at both times in year 8 (waves 1 and 2) and scored ≥12 at year 9. Participants scoring ≥12 at both times in year 8 were classified as having “recurrent” self reported symptoms of anxiety or depression.
Indicators of perceived availability of attachments and conflictual relationships were adapted from the interview schedule for social interaction.16 Perceived availability of attachments was assessed in terms of “having someone to talk to or depend on when angry or upset” or “when having a tough time” and “having someone who knows one well and can be trusted with private feelings and thoughts.” Participants were categorised as having good availability of attachments at both times in year 8, poor availability reported at either time in year 8, or absent or very poor availability at both times in year 8. The social attachment scale has an internal consistency of 0.69.
For conflictual relationships, participants were categorised as reporting no arguments at baseline, arguments with one person at either time in year 8, or arguments with two or more people at either time.
Family measures were family structure (intact family, separated/divorced parents, or other circumstances) and language spoken at home as a marker of ethnicity.
Method of analysis
Results are based on participants for whom information about victimisation and mental health status was available for all waves (2365) or who had missing data at either wave 1 or wave 2 only (194). For these 194 students a conservative assumption was made of no bullying and no symptoms of depression for the wave for which the data were missing.
Simple bivariate associations were estimated by using odds ratios and tested with the χ2 test. To account for the cluster sampling, robust “sandwich” estimates of standard errors were calculated by using survey estimation methods (Stata Statistical Software version 6.0, Stata Corporation, College Station, TX). To model potential confounding effects, multiple logistic regression was used, again with adjustment for clustering using survey estimation methods. Estimates of population attributable fraction adjusted for confounders were made by using logistic regression models with appropriately adjusted 95% confidence intervals.17
Of the sample of 3623 students, 2860 (79%) participated in at least one wave of data collection and 2559 (71%) provided data for this analysis. Small but significant differences were found in some sociodemographic factors for the 222 (8%) with missing data at wave 3, with higher proportions of boys and of students with non-intact families and families of non-English speaking background than among students without missing data.
The prevalence of victimisation at each of the three survey periods was 49% (95% confidence interval 48% to 53%), 51% (49% to 54%), and 42% (39% to 45%). Eight hundred and fifty seven (33%) respondents were defined as having experienced recurrent victimisation, 853 (33%) reported being bullied at one time point, and 849 (33%) reported no victimisation at either time point in year 8; 544 (63%) of those students who were victimised recurrently in year 8 reported being victimised in year 9.
The prevalence of self reported symptoms of anxiety or depression at each of the three survey points was 16% (15% to 18%), 18% (16% to 20%), and 15% (13% to 16%). In all, 1901 (74%) of participants were classified as having no symptoms of anxiety or depression at either wave 1 or wave 2 (clinical interview schedule score <12 at both times), 438 (17%) scored ≥12 on one occasion, and 221 (9%) scored ≥12 on both occasions. Of the 1901 who scored <12 on the schedule in year 8, 136 (7%) scored ≥12 in year 9; 134 (61%) of those with recurrent self reported symptoms of depression in year 8 reported symptoms in year 9.
Simple bivariate analyses found significant associations between victimisation, mental health status, and measures of social relationships (table 1).
The association between incident self reported symptoms of anxiety or depression in year 9 and a history of victimisation in year 8 and the impact of mental health status on the incidence of victimisation, with adjustment for sex, are shown in table 2. Any occurrence of victimisation was significantly associated with the incidence of self reported symptoms of anxiety or depression. After adjustment for social relationships and sociodemographic factors, recurrent victimisation remained significantly associated with incident self reported symptoms of anxiety or depression, as did arguments with others, and sex (table 3).
The attributable fraction of students with incident self reported symptoms of anxiety or depression for those exposed to victimisation was 0.50 (0.24 to 0.67). Adjusted for confounders, the population attributable fraction was 0.30 (0.04 to 0.49). The attributable fraction of students experiencing victimisation for the first time in year 9 who had reported symptoms of anxiety or depression previously was 0.21 (−0.20 to 0.49). Adjusted for confounders, the population attributable fraction was 0.003 (−0.05 to 0.05).
Table 4 shows the adjusted odds ratios for boys and girls, given the known sex differences in self reported symptoms of anxiety or depression and social relationships. For boys, none of the variables remains independently significant in the model. Owing to the small number of incident cases who were boys, this analysis was repeated including only victimisation and arguments with others. This made no substantial difference to the estimation of odds ratios.
The prevalence of victimisation was high and relatively stable in this cohort. Two thirds of the students who were bullied recurrently in year 8 also reported being bullied in year 9. This study confirmed the strong contemporaneous association between victimisation and self reported symptoms of anxiety or depression previously reported. 1 9 18 We also found a strong association with social relationships, which has been less well documented in the adolescent age group. Most importantly, we found that a history of victimisation is a strong predictor of the onset of self reported symptoms of anxiety or depression and remains so after adjustment for other measures of social relations. The contrary hypothesis that having poor emotional health in some way invites victimisation or represents a vicious cycle has not been supported by these data.19-21
Affective disorders become common in adolescence, as symptoms of depression and anxiety increase after puberty. 7 22 A prevalence of 16% of self reported symptoms of anxiety or depression in young secondary school students, with sex differences in the prevalence, is therefore consistent with previous findings. 7 22
In this study, in up to 30% of all students with incident symptoms of depression, the symptoms could be attributed to a history of victimisation, after adjustment for other confounders. Although one must bear in mind the limitations in interpreting population attributable fractions,23 it remains clear that the impact of victimisation on incident self reported symptoms of anxiety or depression in this population is potentially great.
Furthermore, this effect of bullying on mental health status is clearest for girls. That is, being victimised has a significant impact on the future emotional wellbeing of young adolescent girls independent of their social relations but does not for boys. This finding may be due to a real difference in the boys' response to victimisation or to the small number of boys reporting symptoms of depression. However, the second of these possibilities is a less likely explanation, as a reduction of variables in the model did not substantially alter the finding.
The strengths of this study are its prospective design, the use of two time points to define a baseline of recurrent victimisation and self reported symptoms of anxiety or depression, the inclusion of both overt and covert or relational types of victimisation, and a comprehensive measure of mental health status. It is, however, possible that young people who have not previously reported being victimised in year 8, at a time when it is relatively common, may be different from their peers in other respects. Although we cannot explicitly examine this possibility with these data, we believe it to be unlikely given the similar relations of the social and family measures to victimisation and emotional health found in the cross sectional data and in previous studies.
The data were collected as part of the assessment of the effect of a school based intervention.12 The intervention did not contain activities focusing on victimisation, so it is unlikely to have had an impact on the reported associations. Furthermore, all analyses were statistically adjusted for intervention and control status.
This study has found that victimisation raised levels of subsequent self reported symptoms of anxiety or depression regardless of the coexisting levels of social adversity. This suggests that a reduction in victimisation in schools is potentially a useful preventive intervention, especially for girls. Further work is needed to determine if a reduction in victimisation can reduce the onset of symptoms of anxiety and depression in young adolescents, but the indications from this study are that such a reduction could have a substantial impact on the emotional wellbeing of young people.
We acknowledge the valuable contribution made to this study by the staff, young people, and parents in the project schools.
Contributors: LB and GP participated in the design and execution of the study, analysis of the data, and writing the paper. JBC and LT participated in analysing and interpreting the data and writing the paper. KR contributed to the analysis of the cross sectional data and development of the bullying severity scale. LB will act as guarantor for the paper.
Funding The Gatehouse Project is supported by grants from the Queen's Trust for Young Australians, Victorian Health Promotion Foundation, National Health and Medical Research Council, and Department of Human Services.
Competing interests None declared.
Questionnaires used in the study are on the BMJ's website